A method and device for real-time switching of panoramic image viewing angles

Abstract

The invention discloses a method and a device for switching the angle of view of a panoramic image in real time. The method includes: acquiring a plurality of images to be spliced ​​through multiple wide-angle lenses; acquiring a vector sliding by a user through a touch screen; registering the part of the image to be spliced ​​according to the vector; and translating the registered image. For the images to be stitched obtained by multiple wide-angle lenses, consider the situation of sliding during stitching, and realize translation registration while stitching registration. Compared with translation after stitching, the image processing speed can be significantly improved, and the viewing angle can be switched in real time. requirements; and can solve the problem of fuzzy borders of splicing parts through registration, and achieve smooth display effects when users swipe to view.

Description

technical field [0001] The present invention relates to panoramic image processing technology, in particular to a method and device for real-time switching of panoramic image viewing angles. Background technique [0002] The existing technology can use more than two fisheye lenses to collect images, and then stitch them together to obtain a 360-degree panoramic image, which is widely used in wide-angle photography, virtual reality, driving assistance and other scenarios. When we need to view a panoramic image on a mobile device (such as a mobile phone, a car rearview mirror), due to the limitations of the resolution, screen size and viewing angle of the mobile device, one screen can only display a part of the panoramic image. Sometimes it is necessary to change the viewing angle by sliding left and right or up and down until a 360-degree panoramic image is displayed. [0003] In some scenarios, the panoramic image must be updated and displayed in real time, which requires t...

AI Technical Summary

Problems solved by technology

In the prior art, the commonly used method is to display the image of one screen, and then switch to the image of the next screen after sliding, there will be obvious traces of switching in the middle, and smooth switching cannot be achieved; Load an image with a fixed viewing angle, and the user slides to view it on this basis, and the image does not update during the sliding process

These methods require a large amount of image calculation and have high requirements on the computing platform, which will increase the cost of the product; and the image is not smooth enough and easy to blur when sliding to view, and ghosting is prone to appear in the sub-pixel precision part of the splicing boundary, etc.

Images